Abstract
BACKGROUND: Evidence suggests that increased ozone (O(3)) pollution is linked to increased incidence and mortality of cardiovascular disease in general populations. However, the relationship between long-term exposure to ambient O(3) and mortality in stroke patients remains inconclusive. The objective of this study was to investigate the impacts of long-term O(3) exposure on all-cause and cause-specific mortality following first-incident hemorrhagic or ischemic stroke. METHODS: We conducted a retrospective study (from 2013 to 2019) in patients with first-incident stroke. The participants (40 to 99 years of age) were collected from 20 counties in Shandong Province, China. The annual and warm-season average concentrations of O(3) were derived from the daily maximum 8-hour average O(3) data obtained from a published dataset. The associations were determined using Cox-equivalent re-parameterized Poisson model. RESULTS: The current study analyzed 330,486 participants (mean age 68.3 ± 11.6; 46.4% females), including 53,314 hemorrhagic and 277,172 ischemic stroke cases. In hemorrhagic stroke patients, both annual and warm-season O(3) were positively associated with all-cause and stroke-specific mortality, independent of PM(2.5) (particles less than 2.5 μm) and NO(2) (hazard ratios ranging from 1.236 to 1.309, all P < 0.05). Exposure-response analysis indicated a threshold of O(3) concentration (around 120–140 µg/m(3)) for its deleterious effects. The impacts of O(3) on stroke-specific and all-cause mortality were confirmed in subgroup (sex, age, residential areas) and sensitivity (using time-dependent Cox regression and 1-year-lag models) analyses. In contrast, the associations between O(3) and mortality risk in ischemic stroke patients were weak and unstable, which should be interpreted with caution. CONCLUSIONS: Long-term exposure to O(3) pollution is associated with increased mortality in hemorrhagic stroke patients. It appears that O(3) has differential impacts on hemorrhagic and ischemic stroke mortality, although more robust data are needed for ischemic stroke. Future prospective studies are warranted with hemorrhagic and ischemic stroke being analyzed separately. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12889-026-26330-7.